🧹 Data Wrangling Labs – Car & Laptop Datasets

This repository contains my data wrangling labs, where I worked with real-world datasets (cars and laptops) and applied various preprocessing techniques to clean, standardize, and prepare the data for exploratory analysis and machine learning models.

---

📂 Contents

• Car Data_Analysis.ipynb → Data wrangling and analysis on automobile dataset.
• Laptop Data_Analysis.ipynb → Data wrangling and analysis on laptop dataset.
• Laptop_DataSet.csv → Cleaned version of the Laptop dataset.
• cleaned dataset.csv → Cleaned version of the automobile dataset after applying transformations.
• README.md → Project documentation.

---

📊 Dataset Descriptions

🚗 Car Dataset

• Notebook: Car Data_Analysis.ipynb
• Cleaned version saved as: cleaned dataset.csv
• Attributes include: make, fuel-type, aspiration, horsepower, price, etc.
• Use case: Useful for predicting car prices, analyzing market segments, and preparing data for regression/classification models.

💻 Laptop Dataset

• Notebook: Laptop Data_Analysis.ipynb
• Raw dataset file: Laptop_DataSet.csv
• Cleaned version saved as: cleaned dataset.csv
• Attributes include: Brand, Model, Screen_Size_cm, Weight_kg, Processor, RAM, Price, etc.
• Use case: Standardizing laptop specifications, comparing products globally, and preparing features for recommendation or prediction systems.

---

🛠 Techniques Applied

1. Handling Missing Data

• What I did:
  • Detected missing values.
  • Replaced them with mean/median (for numeric data) or most frequent values (for categorical data).
• Why:
  • Missing data causes errors in analysis & models.
  • Ensures dataset is consistent and usable for training ML models.

---

2. Data Type Conversion

• What I did:
  • Converted price and horsepower from object → float.
  • Ensured categorical columns like fuel-type were stored as category.
• Why:
  • Correct data types are critical for numeric operations and memory efficiency.
  • Prevents errors during visualization and ML preprocessing.

---

3. Standardization & Unit Conversion

• Laptop dataset:
  • Converted weight from kg → pounds.
  • Converted screen size from cm → inches.
• Why:
  • Makes data consistent across sources (e.g., some countries use inches, some use cm).
  • Helps comparison of products in global datasets.

---

4. Normalization

• Car dataset:
  • Normalized numerical features like length, width, and height to a range [0,1].
• Why:
  • Brings all numerical features to the same scale.
  • Prevents bias in ML models where features with larger values dominate.

---

5. Binning

• Car dataset:
  • Grouped horsepower into bins: Low, Medium, High.
• Why:
  • Simplifies continuous data into categories for analysis.
  • Makes it easier to study customer segments (e.g., economy vs. sports cars).

---

6. Dummy Variables (One-Hot Encoding)

• Car dataset:
  • Converted categorical variables (fuel-type, aspiration, etc.) into binary columns.
• Why:
  • Many ML algorithms cannot work directly with text categories.
  • Encoding ensures categorical features are machine-readable.

---

7. Feature Engineering

• Laptop dataset:
  • Created new features like Screen_Size_inch and Weight_pounds.
• Why:
  • Derived attributes provide more meaningful insights.
  • Helps in comparing laptops and feeding standardized features into models.

---

🚀 Applications of These Techniques in Real-World Projects

• Car dataset techniques:

  • Price prediction models.
  • Customer segmentation (low/mid/high horsepower).
  • Market trend analysis for automobile companies.

• Laptop dataset techniques:

  • Product comparison engines.
  • Standardized specifications for recommendation systems.
  • Better data integration across global e-commerce platforms.

---

🧰 Tools & Libraries

• Python 3.x
• Jupyter Notebook
• Pandas
• NumPy
• Matplotlib

---

🔑 Learning Outcomes

Through these labs, I gained hands-on experience with:

• Data cleaning (missing values, incorrect types).
• Standardization and normalization.
• Binning and categorical encoding.
• Feature engineering.
• Preparing datasets for visualization and machine learning.

---

📌 Notes

• Car dataset is analyzed in Car Data_Analysis.ipynb and cleaned version is stored in cleaned dataset.csv.
• Laptop dataset is analyzed in Laptop Data_Analysis.ipynb, and cleaned version is also stored in cleaned dataset.csv.
• This work is part of my Data Science coursework on Coursera.